Field evaluation of two mosquito traps in Zhejiang Province, China

Mosquito-borne Diseases are a common but severe public health threat. However, there is a lack of consensus on the effect of different mosquito trapping devices in China. This study firstly compared the BGM trap with the CDC light trap, commonly used in Chinese mosquito surveillance. Field trials of traps' efficiency were conducted in Yiwu city, China, from May 21st, 2018 to November 31st, 2018. Sixty-five comparisons were completed in five different biotopes (an urban residential area, a rural residential area, a park, a hospital, and a pig shelter). Concerning the number of mosquitoes per trap, the BGM trap outperformed three out of five biotopes. In contrast, the CDC light trap only showed better performance in the pig shelter. For specific species, the BGM trap outperformed in capturing Ae. albopictus, while the CDC light trap caught significantly more Cx. tritaeniorhynchus. Regarding Ae. albopictus and Cx. pipiens s.l. surveillance, the BGM trap is more suitable. The BGM trap shows significantly higher or similar efficiency than the CDC light trap in trapping common mosquito species in China, except in the pig shelter. Therefore, we recommend that Chinese researchers and public health practitioners use the BGM trap in future mosquito surveillance.

www.nature.com/scientificreports/ BGM trap is an adaption of BGS trap, which keeps the same mechanism but is more durable 17 . Therefore, BGM is supposedly suitable for long-term mosquito surveillance. However, there is no adequate evidence in China showing BGM's effectiveness. Our study aims to examine the performance of the BGM and the CDC light trap on common mosquito species in China and determine whether BGM can replace CDC light traps in routine mosquito surveillance.

Study site.
Field trials were carried out in Yiwu city, Zhejiang Province, China (119° 49′-120° 17′ E, 29° 02′ 13″ 29° 33′ 40″ N, 56 m above sea level, m.a.s.l.) (Fig. 1). In 2005, the Chinese national disease vector surveillance system (CNDVSS) required that mosquito surveillance tools should be used to monitor five different biotopesurban residential areas, rural residential areas, hospitals, parks, and animal shelters 18 . Based on the guidance, we choose five different locations within Yiwu city to represent five different kinds of biotopes. Beilei (120° 03′ E, 29° 16′ 55″ N) is a pig farm and hence represents the animal shelter; Quanbei (120° 08′ E, 29° 39′ 16″ N) where is a village represents the rural residential area; Danguiyuan (120° 06′ E, 29° 32′ 48″ N) is an urban neighborhood in Yiwu city's downtown and therefore represents the urban residential area; Children's park (120° 07′ E, 29° 29′ 83″ N) is selected as the biotope of park and Choucheng Hospital (120° 09′ E, 29° 33′ 81″ N) is selected as the biotope of hospital (The rural residential zone and the urban residential zone are abbreviated as Urban and Rural, respectively, in this article's charts.). The least distance between any pair of biotopes is 3.44 km (Table 1). Yiwu city has a typical hilly landform located in the subtropical monsoon climate zone, with an average annual temperature of 17 °C and annual precipitation of 1100-1600 mm 19 . The local climate is ideal for mosquito breeding, and adult mosquitoes are active from April to November each year 20 . Known as the largest wholesale market

Results
A total of 13 sampling periods were completed across the study, with 65 trap comparisons evenly conducted in 5 different biotopes. We, with two different-type traps, collected 7406 mosquitoes in total during the study process, including 4636 Cx. pipiens s.l., 1183 Cx. tritaeniorhynchus, 1511 Ae. albopictus, 26 An. sinensis, and 50 Ar. subalbatus (Table 2). Both traps caught all five species by the end of this study, but the proportion of different mosquito species caught varied by traps ( Table 2). The overall female-to-male sex ratios of individuals caught in the BGM trap, and CDC light trap, were both 4:1. Only for the Cx. pipiens s.l., the female-to-male sex ratio of the BGM trap was significantly higher than the CDC light trap (Table 2). During each trapping period, the BGM trap caught significantly more mosquito individuals than the CDC trap, in an urban residential area, park, and hospital (Table 3, Fig. 2). In contrast to comparisons in the other four biotopes, the CDC trap has a significantly better performance in the pig shelter (Table 3, Fig. 2). The performance of the BGM trap and CDC trap in catching mosquitoes varies by mosquito species (Table 4a,b, Fig. 3). In terms of capturing Cx. pipiens s.l., An. Sinensis, Ar. subalbatus, no significant difference between the BGM trap and the CDC trap was observed. As for Ae. albopictus, the BGM trap caught significantly more individuals than the CDC trap per trapping period (p < 0.001). By contrast, for Cx. tritaeniorhynchus, the CDC trap caught significantly more individuals each time (p = 0.0188). Due to the fact that adult male mosquitos do not bite humans, mosquito-borne disease transmission should be mainly attributed to adult female mosquitos. Therefore, our study separated the female and analyzed BGM and CDC traps' effect on trapping female mosquitoes specifically. Results show that the BGM trap's overall performance across all biotopes is better in trapping Ae. albopictus than the CDC trap (Table 4b).
Also, even for the same species, the CDC trap and the BGM trap's performance can vary by biotopes (Fig. 4,  Table 5a,b). Of note, for capturing Ae. albopictus (both sexes and female only), the BGM trap had a significantly better performance in all biotopes except the pig shelter. The population changes of Cx. pipiens s.l. and Ae. albopictus over time was shown in Fig. 5. For Cx. pipiens s.l., the number of captures by BGM traps does not significantly differ that by CDC traps, while more mosquitoes distribute in the hospital, the rural residential area, and the Pig shelter (Z = 5.258, 3.620, 2.484; p = 1.46e−07, 0.0003, 0.013, Table 6). Also, two traps depicted similar trends of population changes of Cx. pipiens s.l. (Fig. 5). For Ae. albopictus, the capture number of BGM traps is significantly more than the number of CDC traps (Z = −4.587; p = 4.50e−06; Table 6). Meanwhile, Ae. albopictus were less likely to be trapped in the pig shelter than the other four places (Z = −1.7575; p = 0.00128; Table 6, Fig. 5). The population change of Ae. albopictus is more conspicuous in BGM traps than in CDC traps (Fig. 5).  37 found that the BGS is more effective than the CDC light trap in sampling adult Ae. albopictus (Z = −25.13, p < 0.001).
Since the BGM trap has the same mechanism as the BGS trap, the performance of the BGM trap in China needs to be studied. To now, only one research in China conducted a field evaluation of BGM traps and humanbaited double net traps in trapping Ae. Albopictus, and the results show BGM traps are more efficient than the human-baited double net traps (t = 2.786, p = 0.006) 38 .
In our study, the different results depending on biotope could be partially explained by the difference in the two traps' working mechanism. The BGM trap captures mosquitoes by emitting CO 2 with human skin odor, while the CDC trap uses light to attract mosquitoes. Pig shelters are usually dark and full of smell, and hogs within the shelter will produce a massive amount of CO 2 through metabolism. In this environment, the BGM trap's odor can be largely diluted, and the gas would conceal CO 2 emission from the trap within the pig shelter. In contrast, the CDC trap's light can easily catch mosquitoes' attention in the dark environment. Thus, it seems clear that CDC light trap should be used to collect mosquitoes in pig shelters, while the BGM trap can be used in other biotopes where the trap's odor and CO 2 emission will not be interrupted.     37 . Our further analysis found that in all biotopes except pig shelter, BGM traps all outperformed CDC traps in terms of amount of Ae. albopictus trapped. The exception could still be explained by pig shelter's unique characteristics, which impeded the BGM trap mechanism. Moreover, our finding that the number of individuals caught differs by sex is similar to other field evaluation' findings in Guangdong Province 37 . This phenomenon could be explained by the biological difference between female and male mosquitoes that the human-skin odor is more attractive to females than males. The male mosquitoes usually stay in the wild and eat nectar, rather than suck human blood in areas where humans live 39 .
In our study, the overall performance of the BGM trap is not significantly different from that of the CDC trap in trapping Cx. pipiens s.l. (p = 0.5623) ( Table 3). This result is inconsistent with a Germany study that the BG trap is superior to any other traps, including CDC light traps, in capturing Cx. pipiens s.l., but consistent with a Spanish study and a Dutch study 30,40,41 . The performance of traps varies across biotopes. We found the same situation that the CDC trap outperformed in the pig shelter environment. However, the BGM traps inversely outperformed in the hospital environment. The former phenomenon can still be attributed to the mechanism www.nature.com/scientificreports/ difference between two traps, while the latter mechanism remains unclear. We researchers speculated that numerous lighting devices in the hospital would keep turning on even at midnight, which indirectly helps BGM traps intrigue mosquitoes.
To represent the population dynamics of the most common mosquito species in Yiwu city (Cx. pipiens s.l. and Ae. albopictus), our study used both BGM traps and CDC traps to monitor the populations' fluctuation of these two mosquito species in five biotopes (Fig. 5, Table 6). Our study shows that BGM traps were more sensitive to the population change of Ae. albopictus, except in the pig shelter; while BGM traps can depict better, or at least equivalent, fluctuations of Cx. pipiens s.l. numbers than CDC traps do, except in the pig shelter. All these findings suggest the BGM trap is suitable for mosquito surveillance in China.
The price of BG-Mosquitaire CO 2 Bundle, consisting of a BGM trap, a BG-Sweetscent (lasting for 2 months), and a BG-Booster CO 2, is $ 279.00 42 . Since the Bundle does not include a CO 2 gas cylinder which costs about $70.00 for 5 lb., the total price of operating a BGM trap is $350.00. By contrast, the price of a domestic CDC light trap is only $30.00. Due to the high cost and relatively late development, the BGM trap has not been ever used in Chinese mosquito surveillance until now. However, our study has shown that the BGM trap is more effective than the CDC light trap in collecting Ae. albopictus, the recognized vector of Dengue virus, therefore should www.nature.com/scientificreports/ can be used in the dengue surveillance. Furthermore, for the other four common mosquito species in China, no significant difference in effectiveness between the BGM trap and CDC trap was detected in four out of the five typical biotopes (except in Pig shelter). Given the limited involvement of pig shelters, compared to urban/rural residential areas, parks, and hospitals, we can say the BGM trap is more suitable than the CDC light trap to be used in daily mosquito surveillance. Our study has several limitations: (1) Our study did not set replicates for the field evaluation in five biotopes over the study periods due to limited human resources. Thus, the representativeness of mosquito distribution in five biotopes and the and external validity of the effectiveness of mosquito trapping devices in each biotope was decreased. (2) Our study aims to compare the effectiveness of mosquito trapping methods commonly used in Chinese mosquito surveillance to BGM traps. In other words, we simulate the operation by not adding CO 2 bait to CDC light traps. However, in this study, to maximize BGM traps' function, we added both BG-Sweetescent bait and CO 2 bait to BGM traps. The difference in equipped bait could potentially influence the results by intrinsically larger attractant effect of BGM traps. Thus, future study should also include a group of CDC light traps with CO 2 bait as a comparison.  www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.